Subj : docs/v321_new.txt src/sbbs3/answer.cpp src/sbbs3/ctrl/MainFormUnit.cpp
To   : Git commit to main/sbbs/master
From : Rob Swindell (on Windows 11)
Date : Sun Jan 11 2026 01:52 am

https://gitlab.synchro.net/main/sbbs/-/commit/418e4413718b7d5ffde113f7
Modified Files:
	docs/v321_new.txt src/sbbs3/answer.cpp src/sbbs3/ctrl/MainFormUnit.cpp src/sbbs3/ntsvcs.c sbbs_ini.c sbbs_ini.h sbbscon.c src/sbbs3/scfg/scfg.c scfgsrvr.c src/sbbs3/startup.h
Log Message:
Built-in detection / short inactivity timeout of dumb terminal (bot) logins

Configured in SCFG->Servers->Terminal Server->Max Dumb Login Inactivity, the
default is one minute (60s). This solution is much more reliable than the
login.js method since it's quite possible that the display of the "answer"
file can pause (e.g. with a [Hit a key] prompt) before the login module even
executes.

Any auto-terminal-type detection (e.g. ANSI, PETSCII, UTF-8) will defeat this
(short) inactivity setting and use the normal login inactivity duration
(now called "Max User Login Inactivity").

The other big change included here is for graceful detection of startup
structure version changes (via size checking). When mixing revisions of SBBS
libs and apps, it was possible to have an insufficiently allocated startup
structure passed around and memory-initialized beyond its bounds or otherwise
have some startup structure elements (e.g. settings) read or written at their
incorrect offset. I discovered this while adding to the bbs_startup_t but
upon review found that there was no size checking for the global_startup_t and
the sbbs_read_ini() function returned no success or failure indication.

.